workaboutcontact

An AI & Machine Learning Digital Twin for the Worlds Wastewater Treatment Centres

Transforming Wastewater Management Through User-Centered Design

The digital wastewater management twin is an AI-enabled platform that replicates the operations of wastewater treatment plants. It empowers operators with real-time data, predictive analytics, and optimization tools to enhance efficiency and sustainability.

Company

HATCH
 

Industry

Wastewater Management

Responsibilities

  • User Experience Lead
  • User Research
  • Creative Direction
  • UI Design
  • LLM Design System

Context

Wastewater treatment operators struggled with manual workflows, unpredictable performance metrics, and compliance pressures. Traditional tools offered limited real-time data or predictive guidance, leaving operators uncertain and disengaged.

Opportunity

We set out to design a digital twin platform that leveraged AI to provide real-time monitoring, predictive maintenance, and transparent recommendations. Our mission was to streamline operations, reduce energy and chemical use, enhance compliance, and uplift operator confidence.

Outcome

The final solution improved operational efficiency by up to 30%, strengthened compliance efforts, and empowered operators with clear, trust-building insights.

Understanding the Problem

Challenges included:

  • Inefficient resource use, leading to higher costs and environmental compliance risks.
  • Operators relying on guesswork due to a lack of timely, actionable data.
  • Unclear shift handovers and limited collaboration tools, reducing team cohesion.

User Implications: Operators felt disempowered, unsure if their decisions were correct or valued. The absence of transparent, data-driven support meant stress, lower morale, and inconsistent results.

Research & Insights

Our Approach:

  • User Interviews & JTBD Analysis: Direct conversations and structured frameworks helped identify operators’ core motivations—not just the tasks they performed, but the progress they aimed to achieve in their roles.
  • On-Site Observations: Watching shift handovers and real-time adjustments revealed critical communication gaps and the need for intuitive dashboards.
  • Surveys & Data Analysis: Validated preferences for explainable AI solutions and clarified which metrics and features mattered most.

Key Findings

1

Operators needed real-time, reliable data to make informed decisions quickly.

2

Emotional reassurance was as important as functional utility—users sought greater confidence in their actions.

3

Social credibility among team members depended on transparent, user-friendly tools that facilitated knowledge sharing.

 

The Design Process

Ideation & Prototyping:

  • Created personas and journey maps rooted in JTBD insights to highlight what operators truly valued.
  • Iterated through wireframes and prototypes, incorporating direct operator feedback to ensure intuitive navigation and actionable analytics.
  • Prioritized explainable AI elements—visualizing the rationale behind recommendations—so operators felt in control and supported, not replaced

Collaboration:

  • Partnered closely with engineers, data scientists, and operations managers to balance technical feasibility with user desirability.
  • Maintained regular feedback loops with operators to refine the interface, ensuring it addressed both their performance goals and comfort levels.

Solution Implementation

LLM Design – Design System – UI Designs

Explainable AI Insights

Customizable dashboards showing live performance metrics, enabling immediate responses to issues.

Scenarios

The system uses AI to generate custom operational scenarios, enabling teams to experiment with different process configurations without risking live plant operations. By adjusting parameters in a virtual environment, operators can identify the most effective strategies and optimize overall performance.

LLM Interface & Interaction

  • Chatbot & Alerts: Operators engage with a built-in chatbot for real-time data and insights, while automated alerts highlight anomalies.
  • Visual Cues: Interactive notifications and scenario builders help users quickly identify issues or opportunities for improvement.
  • Voice Commands: Hands-free control allows staff to query the system or respond to alerts without manual input.

Real-Time Monitoring

Customizable dashboards showing live performance metrics, enabling immediate responses to issues.

Recommendations

Advanced analytics power real-time recommendations that adjust to changing plant conditions. Operators receive clear, context-aware guidance on critical variables—like chemical dosage or operating schedules—ensuring each decision is both data-driven and aligned with efficiency goals.

Testing & Iterataion

Usability Testing:

  • Pilots in live environments let operators trial the platform’s features.
  • Feedback on clarity, navigation, and AI transparency guided iterative refinements

Refinements:

  • Adjusted UI elements for faster data interpretation.
  • Improved explainable AI modules, enhancing trust and adoption.
  • Optimized predictive maintenance features for real-world operator routines.

Results & Impact

Quantitative Gains:

  • Up to 30% reduction in energy and chemical use, cutting costs and environmental footprint.
  • Reduced downtime via proactive maintenance, improving overall plant performance.

Refinements:

  • Operators reported feeling more confident, respected, and better equipped to tackle their roles.
  • Team collaboration improved—shift handovers were smoother, and peer credibility increased.

Challenges & Lessons

Trust in AI:

  • Initial skepticism about automation underscored the need for transparent, explainable insights. This lesson shaped how we communicated value and control in the UI.

User-Centered Focus:

  • Centering design decisions on JTBD principles ensured the solution met real user needs, not just corporate goals.
  • The result: a tool that users voluntarily embraced, not something mandated from above.

Conclusion & Future Steps

By deeply understanding operators’ motivations and concerns, we designed a digital wastewater management twin that delivered tangible efficiency gains, regulatory compliance, and, most importantly, renewed confidence and trust. Looking ahead, expanding predictive capabilities, mobile accessibility, and embedded training can further enhance operator engagement and performance.

This case study demonstrates the power of a user-centered approach combined with advanced technologies. By focusing on what truly matters to the end-user, we created a solution that not only improved bottom-line outcomes but also elevated the human experience in an industrial setting.